The purpose of the proposed research is to develop a reliable and valid method to assess the incidence of putative symptoms of depression in a normal population of prepubertal children. This goal has been conceptualized in two phases. Phase I, the current proposal, deals with the development of a sociometric method--a peer nomination inventory. Phase II, deals with the application of this method, epidemiologically. The subjects for Phase I will be obtained from a pool of approximately 2,000 fourth and fifth grade children in 66 classes in 10 schools in the New York City School System. Symptoms culled from the clinical literature on childhood depression will be translated into items and subjected to the judgement of experts for relevancy to depression. These items, along with other measures, will be administered to classes of children. Iterative item analyses will be performed until a set of items possessing acceptable reliability is demarcated. Subsequently, temporal stability will be studied. Content, concurrent, and construct validity of this peer nomination inventory will be investigated by correlation with self-ratings, data from a parent interview, school record data, teacher ratings, and census tract data.